Network Position Affects Social Status in Early Adolescence

ABSTRACT

Does position in the friendship network affect social status in early adolescence? Previous research shows that friendship relations are not equivalent to popularity hierarchies, but favorable positions in the friendship network should allow students to gain status. We use four waves of longitudinal network data and dynamic panel models with fixed effects to estimate the impact of network position on social status. Degree centrality and brokerage opportunity both exert large impacts on status, even compared against the effect of prior status. The results suggest future research on friendship across social identities, as well as network structure and hierarchy in adolescent social systems.

KEYWORDS:

• Centrality
• brokerage
• networks
• status
• popularity

Acknowledgments

We thank three anonymous reviewers for their comments on this manuscript, and Kerrylin Lambert for assistance with the data.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. This beta weight was selected because it was the largest weight where the algorithm returned values for every member of the network. Larger beta weights are more appropriate for this analysis than smaller ones to distinguish it from degree centrality.

2. Betweenness centrality calculates brokerage based on the number of times a node lies on the shortest path between two alters (Freeman 1978). Unfortunately, it does not account for path length, which means that in larger networks it includes paths between alters that are unlikely to ever be exercised. If a node lies on the shortest path between two alters, but the shortest path is still a multi-person chain that requires negotiations at each step, it is rare that there will be an actual brokerage opportunity (Burt 2010).

3. Figures 1–3 are created using lavaanPlot (Lishinski 2018).

4. Ragan et al. (2022) demonstrate that stochastic actor-oriented models are not more conservative than conventional regression approaches, they are subject to omitted variable bias from stable individual-level factors, and furthermore these stable individual-level differences are a greater threat to inference than explicitly modeling friendship selection processes. In contrast, our dynamic panel models address confounding from stable, individual-level differences.

5. In more detail: We calculate the differences between the network measures at waves 1 and 2; at waves 2 and 3; and at waves 3 and 4, and then repeat the same process for status.

Additional information

Funding

This research was supported in part by grants from the Institute of Educational Sciences (R305A120812) awarded to Thomas W. Farmer, Jill V. Hamm, and David Lee (PIs). The views expressed in this article are those of the authors and do not necessarily reflect the views of the granting agency.

Notes on contributors

Jonathan Horowitz

Jonathan Horowitz is an Assistant Professor of Sociology at the University of Toronto. He studies effect of educational institutions on position in the labor market and communities, and the resulting effects on inequality and political participation.

Jill Hamm

Jill Hamm is the William C. Friday Distinguished Professor of Education at the University of North Carolina-Chapel Hill. Hamm’s primary research interest is the role of peer relationships in adolescents’ adjustment.